Experiments on human cancer cells in vitro have been completed and demonstrate that certain anticancer drug-induced changes are directly correlated with, and predictive of the cell killing effects of the same drug. These results form the basis for the experiments described in this proposal. The specific goal of this work is to provide a rapid and accurate cell-based assay for anticancer drug sensitivity or resistance that can be made directly on the patient's own cancer and normal cells. Drugs specific for breast, lung or gastrointestinal cancers will be used to treat tissue fragments (or single cells). Assays performed 24 hours after the end of treatment will be used to detect changes in cancer and normal cells. These effects will be correlated with the intracellular levels of resistance modulators. The test to determine the prognostic value of the chemosensitivity assay will be made by prospective evaluation of that patient's clinical response to treatment. Although the assay is designed to predict resistance to treatment, the more important aspect of the study is that the results will also identify which drugs will be most effective in the treatment of that patient's cancer.